A growing body of evidence supports the view that regulatory evolution - the evolution of where and when a gene is expressed - is the primary genetic mechanism behind the modular organization, functional diversification, and origin of novel traits in higher organisms. Most elements regulating gene expression in eukaryotic genomes reside in noncoding DNA (i.e. DNA that does not encode protein). Recent studies suggest that much of the noncoding portion of the Drosophila melanogaster genome is evolutionarily constrained, implying that these regions are important for an organism<s fitness and may be the target of substantial adaptive evolution. We propose to use a combination of novel computational and experimental approaches to 1) identify cis-regulatory untranslated transcribed regions (UTRs) that may have been targets of recurrent adaptive evolution and 2) experimentally test the effects of putatively functional substitutions on levels of gene expression divergence between species. We will begin by collecting population genomic variability data from 25 naturally occurring strains of D. simulans for all 5<and 3<UTRs with full length transcripts (~2.2Mb) and a control reference panel of closely-linked short introns and coding sequence (~5.8Mb). We will use and further develop computational methods to identify UTRs that have accumulated adaptive sequence divergence between species using population genetic data of this kind. Specifically, we will explore to what extent using the allelic frequency spectrum and integrating this new data with emerging population genomic data for D. melanogaster can improve the fidelity of population genetic tests for selection. We will then use D. melanogaster as an experimental model to functionally verify predictions based on computational methods. Specifically, we will use a transgene co-placement method to determine the effects of 3<UTR divergence on gene expression divergence and test alternative hypotheses about how individual functional substitutions interact and contribute to changes in gene expression. These experiments will, in turn, be used to refine our computational prediction methods. This research will identify new cis- regulatory elements, develop novel methodologies for mapping such elements and provide important insights into how gene regulatory changes have led to the evolution of new species and diversity in animal forms. The computational methods and biological intuitions we develop will be widely applicable to other model systems, including humans. PUBLIC HEALTH RELEVANCE: Changes in genetic regulation contribute to adaptations in natural populations and influence susceptibility to human diseases (Gilad et al. 2008;Gobbi et al. 2006). Despite their potential phenotypic importance, the selective pressures acting on regulatory processes and gene expression levels in particular are largely unknown. Our research combines computational and experimental approaches to study how natural selection acts on genetic variation underlying both beneficial and detrimental functional differences in gene expression. This work will significantly improve our understanding of biology of human diseases caused by the misexpression of genes.